RESUMO
Structured Illumination Microscopy (SIM) is a widespread methodology to image live and fixed biological structures smaller than the diffraction limits of conventional optical microscopy. Using recent advances in image up-scaling through deep learning models, we demonstrate a method to reconstruct 3D SIM image stacks with twice the axial resolution attainable through conventional SIM reconstructions. We further demonstrate our method is robust to noise and evaluate it against two-point cases and axial gratings. Finally, we discuss potential adaptions of the method to further improve resolution. This article is part of the Theo Murphy meeting issue 'Super-resolution structured illumination microscopy (part 1)'.
Assuntos
Aprendizado Profundo , Microscopia de Fluorescência/métodos , Animais , Cromatina/ultraestrutura , Simulação por Computador , Humanos , Processamento de Imagem Assistida por Computador/métodos , Processamento de Imagem Assistida por Computador/estatística & dados numéricos , Imageamento Tridimensional/métodos , Imageamento Tridimensional/estatística & dados numéricos , Microscopia Confocal/métodos , Microscopia Confocal/estatística & dados numéricos , Microscopia de Fluorescência/estatística & dados numéricos , Fenômenos ÓpticosRESUMO
EBI metagenomics (http://www.ebi.ac.uk/metagenomics) provides a free to use platform for the analysis and archiving of sequence data derived from the microbial populations found in a particular environment. Over the past two years, EBI metagenomics has increased the number of datasets analysed 10-fold. In addition to increased throughput, the underlying analysis pipeline has been overhauled to include both new or updated tools and reference databases. Of particular note is a new workflow for taxonomic assignments that has been extended to include assignments based on both the large and small subunit RNA marker genes and to encompass all cellular micro-organisms. We also describe the addition of metagenomic assembly as a new analysis service. Our pilot studies have produced over 2400 assemblies from datasets in the public domain. From these assemblies, we have produced a searchable, non-redundant protein database of over 50 million sequences. To provide improved access to the data stored within the resource, we have developed a programmatic interface that provides access to the analysis results and associated sample metadata. Finally, we have integrated the results of a series of statistical analyses that provide estimations of diversity and sample comparisons.